Rodney Strachan, "Comment on 'Jointness of Growth Determinants' by Gernot
Doppelhofer and Melvyn Weeks", Journal of Applied Econometrics, Vol. 24,
No. 2, 2009, pp. 245-247.

This file describes the contents of the data file: 'rs-data.txt'. This file,
which is an ASCII file in DOS format, is zipped in the file rs-data.zip.
Unix users should use "unzip -a".

The data file contains the values of the new statistic developed in this
comment. This statistic is plotted in the right panel of Figure 1. The
values of the new statistic are computed from the marginal probabilities of
p(i|y) and p(i|y) and the Jointness statistic in Table 1 of Doppelhofer and
Weeks (DW). Each value of the statistic measures the Jointness of two
variables.

The left panel of Figure 1 plots the values of the Jointness statistic in
Table 1 in the appendix of 'Jointness of Growth Determinants' by Gernot
Doppelhofer and Melvyn Weeks. We therefore do not reproduce these values
here. However, it would be useful for the reader to study the Table 1 in DW
to understand the values in the data file. DW report their jointness
statistic values for all pairs of variables in a 67 x 67 symmetric matrix in
their Table 1. The data file 'rs-data.txt' gives the values of the
jointness statistic from Strachan for all pairs of variables in a 2211 x 1
vector (i.e., 2211 = 67x(67-1)/2).

The variables used by DW to compute their jointness statistic are described
in DW. They assign to each variable an identifying number. We use the same
identifying numbers and list these at the end of this README file.

The data file 'rs-data.txt' contains three columns. The first two identify
the variables in each pairing, and the third column reports the value of the
jointness statistic for each pair. We give further explanation below.

FIRST AND SECOND COLUMNS:

The first two columns of the file 'rs-data.txt' give the identifying numbers
of two variables. Thus each row of the first two columns identifies a pair
of variables. As the jointness statistic is for pairs of variables, the
first column includes variables 2 to 67, and the second column includes
variables 1 to 66. That is, the first two columns give the pairs
      
      (2,1), 
      (3,1), 
        .
        .
        .
      (66,1), 
      (67,1), 
      (3,2), 
      (4,2), 
        .
        .
        .
      (66,2), 
      (67,2), 
      (4,3), 
      (5,3), 
        .
        .
        .
      (66,3),
      (67,3),
        .
        .
        .
 etc. up to the final pairs
 
      (67,65) and 
      (67,66).

These couplets are exactly the same as those in Table 1 in the appendix of
'Jointness of Growth Determinants' by Gernot Doppelhofer and Melvyn Weeks.

THIRD COLUMN:

The third column in the data file reports the new jointness statistic
proposed in the comment. The Jointness statistic in the third column is for
that combination of variables identified by the first two columns. 

As mentioned above, the variables are identified using the same numbering
system as in DW. We reproduce the variable names and identifying numbers
below.

To aid interpretation, I provide a number of examples. 

The first row of the data file is: 
2	1	2.013690622
Using the identifiers below, this gives a jointness statistic for
P60 (variable 2) and EAST (variable 1) of 2.013690622. 

The second row of the data file is: 
3	1	1.878965808
Using the identifiers below, this gives a jointness statistic for
IPRICE1 (variable 3) and EAST (variable 1) of 1.878965808.

The 196th row of the data file is: 
5	4	0.953812588
Using the identifiers below, this gives a jointness statistic for
TROPICAR (variable 5) and GDPCH60L (variable 4) of 0.953812588.

The last row of the data file is: 
67	66	-0.000000164
Using the identifiers below, this gives a jointness statistic for
DENS65I (variable 67) and LANDAREA (variable 66) of 0.953812588.


Number  Variable Name
1	EAST
2	P60
3	IPRICE1
4	GDPCH60L
5	TROPICAR
6	DENS65C
7	MALFAL66
8	LIFE060
9	CONFUC
10	SAFRICA
11	LAAM
12	SPAIN
13	MINING
14	GVR61
15	MUSLIM00
16	YRSOPEN
17	AVELF
18	BUDDHA
19	DENS60
20	RERD
21	OTHFRAC
22	OPENDEC1
23	PRIGHTS
24	GOVSH61
25	H60
26	PRIEXP70
27	TROPPOP
28	GGCFD3
29	PROT00
30	HINDU00
31	POP1560
32	AIRDIST
33	GOVNOM1
34	ABSLATIT
35	CATH00
36	COLONY
37	FERTLDC1
38	CIV72
39	REVCOUP
40	EUROPE
41	SCOUT
42	BRIT
43	LHCPC
44	TOT1DEC1
45	POP60
46	GDE1
47	POP6560
48	NEWSTATE
49	SIZE60
50	ENGFRAC
51	PI6090
52	LANDLOCK
53	LT100CR
54	HERF00
55	DPOP6090
56	OIL
57	GEEREC1
58	SQPI6090
59	SOCIALIST
60	ZTROPICS
61	WARTIME
62	TOTIND
63	WARTORN
64	ORTH00
65	ECORG
66	LANDAREA
67	DENS65I